
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
import seaborn as sns
import sys
import os

from sklearn.dummy import DummyRegressor
from sklearn.tree import DecisionTreeRegressor
from sklearn.ensemble import RandomForestRegressor
from sklearn.linear_model import LinearRegression
from sklearn.model_selection import cross_val_score
from sklearn.metrics import mean_absolute_error, make_scorer

pd.set_option('display.max_columns', None)
pd.set_option('display.max_rows', 100)
np.set_printoptions(threshold=sys.maxsize)

RANDOM_STATE = 1


PATHS = [r'C:\Users\Георгий\Documents\DS_projects', '/datasets', '/content']

def read_df(name, paths=PATHS):
    '''
    Функция загрузки данных из CSV-файлов
    с проверкой существования пути к файлу.

    Принимает название файла.
    Возвращает набор данных из файла.
    '''

    for path in paths:
        file_path = os.path.join(path, name)
        if os.path.exists(file_path):
            df = pd.read_csv(file_path)

            return df

    print(f'{name}: Something is wrong')


train = read_df('gold_industry_train.csv')
test = read_df('gold_industry_test.csv')
full = read_df('gold_industry_full.csv')


data_dict = {'train': train,
             'test': test,
             'full': full}


for name in data_dict:
    display(name)
    display(data_dict[name].head(2))
    display(data_dict[name].tail(2))

'train'

'test'

'full'


data_name = []
data_size = []

for name in data_dict:
    data_name.append(name)
    data_size.append(data_dict[name].shape)

pd.DataFrame(data_size,
             columns=['Количество строк', 'Количество столбцов'],
             index=data_name)


train.shape[0] + test.shape[0]

19439


data_name = []
columns_duplicated = []

for name in data_dict:
    data_name.append(name)
    columns_duplicated.append(data_dict[name].columns.duplicated().sum())

pd.DataFrame(columns_duplicated,
             columns=['Количество дубликатов столбцов'],
             index=data_name)


train_columns = []
test_columns = []
full_columns = []
count_train_columns = 0
count_test_columns = 0
count_full_columns = 0

for column_name in full.columns.sort_values():
    if column_name in train.columns:
        train_columns.append('+')
        count_train_columns +=1
    else:
        train_columns.append('')
    if column_name in test.columns:
        test_columns.append('+')
        count_test_columns +=1
    else:
        test_columns.append('')
    if column_name in full.columns:
        full_columns.append('+')
        count_full_columns +=1
    else:
        full_columns.append('')

pd.concat([
    pd.DataFrame({'train_columns': train_columns,
                  'test_columns': test_columns,
                  'full_columns': full_columns},
                 index = full.columns.sort_values().tolist()),
    pd.DataFrame({'Всего признаков': [count_train_columns,
                                      count_test_columns,
                                      count_full_columns]},
                 index = ['train_columns', 'test_columns', 'full_columns']).T
])


data_name = []
data_size = []

for name in data_dict:
    data_name.append(name)
    data_size.append(data_dict[name].size)

pd.DataFrame(data_size,
             columns=['Количество элементов данных'],
             index=data_name)


data_name = []
data_types = pd.DataFrame()

for name in data_dict:
    data_name.append(name)
    data_types = pd.concat([
        data_types,
        pd.DataFrame(data_dict[name].dtypes.value_counts(), columns=[name]).T
    ])

data_types


total = full.isna().sum().sort_values(ascending=False)
percent = (full.isna().mean() * 100).round(1).sort_values(ascending=False)
pd.concat([total, percent], axis=1, keys=['Всего пропусков', '%'])


def describe_data(columns):
    '''
    Функция рассчитывает краткие описательные статистики
    и строит графики распределений.

    Принимает список с названиями признаков (именами колонок)
    набора данных 'full'.
    Возвращает DataFrame с описательными статистиками,
    диаграмму распределения значений и диаграмму размаха.
    '''

    display(full[columns].describe().round(2).T)

    for column in columns:
        plt.figure(figsize=(5, 3))
        plt.title(label=f'Диаграмма распределения\nзначений признака\n{column}')
        plt.hist(full[column], bins=100)
        plt.xlabel('Значение признака')
        plt.ylabel('Частота')
        plt.show()

    full[columns].plot(kind='box', vert=False, figsize=(10, 5))
    plt.title('Диаграммы размаха значений признаков')
    plt.xlabel('Значение признака')
    plt.ylabel('Признак')
    plt.show()


def freq_table_min(columns):
    '''
    Функция выводит 5 уникальных значений из левого "хвоста" распределения.

    Принимает список с названиями признаков (именами колонок)
    набора данных 'full'.
    Возвращает DataFrame с 5 уникальными минимальными значениями
    в распределении для каждого признака.
    '''

    result = pd.DataFrame()

    for column in columns:
        result = result.join(
            pd.DataFrame(full[column].value_counts().sort_index().head().index,
                         columns=[column]), how='outer')

    return result


def freq_table_max(columns):
    '''
    Функция выводит 5 уникальных значений из правого "хвоста" распределения.

    Принимает список с названиями признаков (именами колонок)
    набора данных 'full'.
    Возвращает DataFrame с 5 уникальными максимальными значениями
    в распределении для каждого признака.
    '''

    result = pd.DataFrame()

    for column in columns:
        result = result.join(
            pd.DataFrame(full[column].value_counts()\
                         .sort_index(ascending=False).head().index,
                         columns=[column]), how='outer')

    return result


len(full['date'].unique())

19439


columns = ['rougher.input.feed_ag',
           'rougher.output.concentrate_ag',
           'primary_cleaner.output.concentrate_ag',
           'final.output.concentrate_ag',
           'rougher.output.tail_ag',
           'primary_cleaner.output.tail_ag',
           'secondary_cleaner.output.tail_ag',
           'final.output.tail_ag']


describe_data(columns)
plt.show()


freq_table_min(columns).T


columns = ['rougher.input.feed_au',
           'rougher.output.concentrate_au',
           'primary_cleaner.output.concentrate_au',
           'final.output.concentrate_au',
           'rougher.output.tail_au',
           'primary_cleaner.output.tail_au',
           'secondary_cleaner.output.tail_au',
           'final.output.tail_au']


describe_data(columns)
plt.show()


freq_table_min(columns).T


columns = ['rougher.input.feed_pb',
           'rougher.output.concentrate_pb',
           'primary_cleaner.output.concentrate_pb',
           'final.output.concentrate_pb',
           'rougher.output.tail_pb',
           'primary_cleaner.output.tail_pb',
           'secondary_cleaner.output.tail_pb',
           'final.output.tail_pb']


describe_data(columns)
plt.show()


freq_table_min(columns).T


columns = ['rougher.input.feed_sol',
           'rougher.output.concentrate_sol',
           'primary_cleaner.output.concentrate_sol',
           'final.output.concentrate_sol',
           'rougher.output.tail_sol',
           'primary_cleaner.output.tail_sol',
           'secondary_cleaner.output.tail_sol',
           'final.output.tail_sol']


describe_data(columns)
plt.show()


freq_table_min(columns).T


columns = ['rougher.input.floatbank10_sulfate',
           'rougher.input.floatbank10_xanthate',
           'rougher.input.floatbank11_sulfate',
           'rougher.input.floatbank11_xanthate',
           'primary_cleaner.input.depressant',
           'primary_cleaner.input.sulfate',
           'primary_cleaner.input.xanthate']


describe_data(columns)
plt.show()


freq_table_min(columns).T


freq_table_max(columns).T


columns = ['rougher.input.feed_rate',
           'rougher.input.feed_size',
           'primary_cleaner.input.feed_size']


describe_data(columns)
plt.show()


freq_table_min(columns).T


columns = ['rougher.state.floatbank10_a_air',
           'rougher.state.floatbank10_b_air',
           'rougher.state.floatbank10_c_air',
           'rougher.state.floatbank10_d_air',
           'rougher.state.floatbank10_e_air',
           'rougher.state.floatbank10_f_air',
           'primary_cleaner.state.floatbank8_a_air',
           'primary_cleaner.state.floatbank8_b_air',
           'primary_cleaner.state.floatbank8_c_air',
           'primary_cleaner.state.floatbank8_d_air']


describe_data(columns)
plt.show()


freq_table_min(columns).T


columns = ['secondary_cleaner.state.floatbank2_a_air',
           'secondary_cleaner.state.floatbank2_b_air',
           'secondary_cleaner.state.floatbank3_a_air',
           'secondary_cleaner.state.floatbank3_b_air',
           'secondary_cleaner.state.floatbank4_a_air',
           'secondary_cleaner.state.floatbank4_b_air',
           'secondary_cleaner.state.floatbank5_a_air',
           'secondary_cleaner.state.floatbank5_b_air',
           'secondary_cleaner.state.floatbank6_a_air']


describe_data(columns)
plt.show()


freq_table_min(columns).T


columns = ['rougher.state.floatbank10_a_level',
           'rougher.state.floatbank10_b_level',
           'rougher.state.floatbank10_c_level',
           'rougher.state.floatbank10_d_level',
           'rougher.state.floatbank10_e_level',
           'rougher.state.floatbank10_f_level',
           'primary_cleaner.state.floatbank8_a_level',
           'primary_cleaner.state.floatbank8_b_level',
           'primary_cleaner.state.floatbank8_c_level',
           'primary_cleaner.state.floatbank8_d_level']


describe_data(columns)
plt.show()


freq_table_max(columns).T


columns = ['secondary_cleaner.state.floatbank2_a_level',
           'secondary_cleaner.state.floatbank2_b_level',
           'secondary_cleaner.state.floatbank3_a_level',
           'secondary_cleaner.state.floatbank3_b_level',
           'secondary_cleaner.state.floatbank4_a_level',
           'secondary_cleaner.state.floatbank4_b_level',
           'secondary_cleaner.state.floatbank5_a_level',
           'secondary_cleaner.state.floatbank5_b_level',
           'secondary_cleaner.state.floatbank6_a_level']


describe_data(columns)
plt.show()


freq_table_max(columns).T


full[['rougher.calculation.au_pb_ratio']].describe().T


plt.figure(figsize=(10, 3))
plt.title('Диаграмма распределения значений признака' +
          '\nrougher.calculation.au_pb_ratio')
plt.hist(full['rougher.calculation.au_pb_ratio'], bins=100)
plt.ylabel('Частота')
plt.xlabel('Значение признака')
plt.show()


full[['rougher.calculation.floatbank10_sulfate_to_au_feed',
      'rougher.calculation.floatbank11_sulfate_to_au_feed']].describe().T


plt.figure(figsize=(10, 3))
plt.title('Диаграмма распределения значений признака' +
          '\nrougher.calculation.floatbank10_sulfate_to_au_feed')
plt.hist(full['rougher.calculation.floatbank10_sulfate_to_au_feed'], bins=100)
plt.ylabel('Частота')
plt.xlabel('Значение признака')
plt.show()


plt.figure(figsize=(10, 3))
plt.title('Диаграмма распределения значений признака' +
          '\nrougher.calculation.floatbank11_sulfate_to_au_feed')
plt.hist(full['rougher.calculation.floatbank11_sulfate_to_au_feed'], bins=100)
plt.ylabel('Частота')
plt.xlabel('Значение признака')
plt.show()


full[['rougher.calculation.sulfate_to_au_concentrate']].describe().T


plt.figure(figsize=(10, 3))
plt.title('Диаграмма распределения значений признака' +
          '\nrougher.calculation.sulfate_to_au_concentrate')
plt.hist(full['rougher.calculation.sulfate_to_au_concentrate'], bins=100)
plt.ylabel('Частота')
plt.xlabel('Значение признака')
plt.show()


full[['rougher.output.recovery',
      'final.output.recovery']].describe().T


fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(10, 5))
ax1.set_title('Диаграмма распределения и диаграмма размаха' +
              '\nзначений признака rougher.output.recovery')
ax1.hist(full['rougher.output.recovery'], bins=100)
ax1.tick_params(labelbottom=False)
ax1.set_ylabel('Частота')
ax2.boxplot(full['rougher.output.recovery'], vert=False)
ax2.tick_params(left=False, labelleft=False)
ax2.set_xlabel('Значение признака')
plt.show()


fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(10, 5))
ax1.set_title('Диаграмма распределения и диаграмма размаха' +
              '\nзначений признака final.output.recovery')
ax1.hist(full['final.output.recovery'], bins=100)
ax1.tick_params(labelbottom=False)
ax1.set_ylabel('Частота')
ax2.boxplot(full['final.output.recovery'], vert=False)
ax2.tick_params(left=False, labelleft=False)
ax2.set_xlabel('Значение признака')
plt.show()


data = pd.DataFrame({
    'auto_estimated_recovery': train['rougher.output.recovery'].tolist(),
    'recalculated_recovery':
    (train['rougher.output.concentrate_au'] *
     (train['rougher.input.feed_au'] - train['rougher.output.tail_au']) /
     (train['rougher.input.feed_au'] *
      (train['rougher.output.concentrate_au'] - train['rougher.output.tail_au'])
      ) * 100)
    .tolist()
})


data.sample(7)


format(mean_absolute_error(data['auto_estimated_recovery'],
                           data['recalculated_recovery']), '.1g')

'1e-14'


for name in data_dict:
    data_dict[name].set_index('date', inplace=True)
    data_dict[name].drop(
        data_dict[name].loc[:, data_dict[name].columns.str.contains('calculation')]
        .columns,
        axis=1,
        inplace=True)


test = full.loc[test.index.tolist()]


test.shape

(4860, 82)


train.shape

(14579, 82)


full.shape

(19439, 82)


test.shape[0] + train.shape[0]

19439


data_dict = {'train': train,
             'test': test,
             'full': full}


for name in data_dict:
    for metall in ['au', 'ag', 'pb', 'sol']:
        for column in data_dict[name].loc[:, data_dict[name].columns.str.contains(metall)].columns:
            data_dict[name].drop(
                data_dict[name].loc[data_dict[name][column] <= 0.01001].index,
                axis=0,
                inplace=True)


test.shape

(4683, 82)


train.shape

(13110, 82)


full.shape

(17793, 82)


test.shape[0] + train.shape[0]

17793


for name in data_dict:
    for subtance in ['sulfate', 'xanthate', 'depressant']:
        for column in data_dict[name].loc[:, data_dict[name].columns.str.contains(subtance)].columns:
            data_dict[name].drop(
                data_dict[name].loc[data_dict[name][column] <= 0.001].index |
                data_dict[name].loc[data_dict[name][column] == 40].index,
                axis=0,
                inplace=True)

<ipython-input-73-58d285c9fc56>:5: FutureWarning: Index.__or__ operating as a set operation is deprecated, in the future this will be a logical operation matching Series.__or__.  Use index.union(other) instead.
  data_dict[name].loc[data_dict[name][column] <= 0.001].index |


test.shape

(4680, 82)


train.shape

(13087, 82)


full.shape

(17767, 82)


test.shape[0] + train.shape[0]

17767


for name in data_dict:
    data_dict[name].drop(
        data_dict[name].loc[data_dict[name]['rougher.input.feed_rate'] < 5].index,
        axis=0,
        inplace=True)


test.shape

(4668, 82)


train.shape

(13063, 82)


full.shape

(17731, 82)


test.shape[0] + train.shape[0]

17731


for name in data_dict:
    data_dict[name].drop(
        data_dict[name].loc[data_dict[name]['rougher.input.feed_size'] < 1].index,
        axis=0,
        inplace=True)


test.shape

(4667, 82)


train.shape

(13060, 82)


full.shape

(17727, 82)


test.shape[0] + train.shape[0]

17727


for name in data_dict:
    for column in data_dict[name].loc[:, data_dict[name].columns.str.contains('air') &
                                      data_dict[name].columns.str.contains('secondary_cleaner.state.floatbank')
                                      ].columns:
        data_dict[name].drop(
            data_dict[name].loc[data_dict[name][column] < 0.1].index,
            axis=0,
            inplace=True)


test.shape

(4626, 82)


train.shape

(12936, 82)


full.shape

(17562, 82)


test.shape[0] + train.shape[0]

17562


for name in data_dict:
    for column in data_dict[name].loc[:, data_dict[name].columns.str.contains('air') &
                                      (data_dict[name].columns.str.contains('primary_cleaner.state.floatbank8') |
                                       data_dict[name].columns.str.contains('rougher.state.floatbank10'))
                                      ].columns:
        data_dict[name].drop(
            data_dict[name].loc[data_dict[name][column] < 100].index,
            axis=0,
            inplace=True)


test.shape

(4626, 82)


train.shape

(12933, 82)


full.shape

(17559, 82)


test.shape[0] + train.shape[0]

17559


for name in data_dict:
    for column in data_dict[name].loc[:, data_dict[name].columns.str.contains('level')].columns:
        data_dict[name].drop(
            data_dict[name].loc[data_dict[name][column] > -100].index,
            axis=0,
            inplace=True)


test.shape

(4626, 82)


train.shape

(12929, 82)


full.shape

(17555, 82)


test.shape[0] + train.shape[0]

17555


for name in data_dict:
    print('Количество пропущенных значения набора данных', name,
          'составляет', data_dict[name].isna().sum().sum())

Количество пропущенных значения набора данных train составляет 2080
Количество пропущенных значения набора данных test составляет 1375
Количество пропущенных значения набора данных full составляет 3455


for name in data_dict:
    data_dict[name].fillna(method='ffill', inplace=True)


for name in data_dict:
    print('Количество пропущенных значения набора данных', name,
          'составляет', data_dict[name].isna().sum().sum())

Количество пропущенных значения набора данных train составляет 0
Количество пропущенных значения набора данных test составляет 0
Количество пропущенных значения набора данных full составляет 0


def descriptive_statistics(values, label, column_name):
    '''
    Описательная статистика переменной, измеряемой в количественной шкале.

    Принимает значения признака, название признака и общее название колонок.
    Возвращает DataFrame с набором статистик.

    '''

    df = pd.DataFrame([
        values.count(),
        len(values.unique()),
        values.min(),
        values.median() - 1.5 * (values.quantile(q=.75) -
                                 values.quantile(q=.25)),
        values.quantile(q=.25).round(2),
        (values.mode()).to_list(),
        values.median().round(2),
        values.mean().round(2),
        values.quantile(q=.75).round(2),
        values.median() + 1.5 * (values.quantile(q=.75) -
                                 values.quantile(q=.25)),
        values.max(),
        values.max() - values.min(),
        values.quantile(q=.75) - values.quantile(q=.25)
    ],
        index=['кол-во значений', 'кол-во уникальных', 'мин.',
               '-1.5IQR', '25 %', 'мода', 'медиана',
               'среднее ариф.', '75 %', '+1.5IQR', 'макс.',
               'размах', 'межквартильный размах'],
        columns=[label]).rename_axis(column_name, axis=1)

    return df


col_name = 'Содержание золота в потоках, усл. ед.'
descriptive_statistics(
    full['rougher.input.feed_au'], 'В сырье', col_name).join(
descriptive_statistics(
    full['rougher.output.concentrate_au'], 'Со стадии флотации', col_name)).join(
descriptive_statistics(
    full['primary_cleaner.output.concentrate_au'], 'Со стадии очистки 1', col_name)).join(
descriptive_statistics(
    full['final.output.concentrate_au'], 'Во флотоконцентрате', col_name))


columns = ['rougher.input.feed_au',
           'rougher.output.concentrate_au',
           'primary_cleaner.output.concentrate_au',
           'final.output.concentrate_au']
label = ['в сырье',
         'со стадии флотации',
         'со стадии очистки 1',
         'во флотоконцентрате']


fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(10, 6))
ax1.set_title('Диаграмма распределения частот и диаграмма размаха' +
              '\nзначений содержания золота в потоках')
for column in columns:
    ax1.hist(full[column], bins=100, alpha=0.5)
ax1.tick_params(labelbottom=False)
ax1.set_ylabel('Частота')
ax1.legend(label)
ax2.boxplot(full[columns], vert=False)
ax2.set_yticks(ticks=[1,2,3,4], labels=label)
ax2.set_xlabel('Содержание золота, усл. ед.')
plt.show()


col_name = 'Содержание серебра в потоках, усл. ед.'
descriptive_statistics(
    full['rougher.input.feed_ag'], 'В сырье', col_name).join(
descriptive_statistics(
    full['rougher.output.concentrate_ag'], 'Со стадии флотации', col_name)).join(
descriptive_statistics(
    full['primary_cleaner.output.concentrate_ag'], 'Со стадии очистки 1', col_name)).join(
descriptive_statistics(
    full['final.output.concentrate_ag'], 'Во флотоконцентрате', col_name))


columns = ['rougher.input.feed_ag',
           'rougher.output.concentrate_ag',
           'primary_cleaner.output.concentrate_ag',
           'final.output.concentrate_ag']
label = ['в сырье',
         'со стадии флотации',
         'со стадии очистки 1',
         'во флотоконцентрате']


fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(10, 6))
ax1.set_title('Диаграмма распределения частот и диаграмма размаха' +
              '\nзначений содержания серебра в потоках')
for column in columns:
    ax1.hist(full[column], bins=100, alpha=0.5)
ax1.tick_params(labelbottom=False)
ax1.set_ylabel('Частота')
ax1.legend(label)
ax2.boxplot(full[columns], vert=False)
ax2.set_yticks(ticks=[1,2,3,4], labels=label)
ax2.set_xlabel('Содержание серебра, усл. ед.')
plt.show()


col_name = 'Содержание свинца в потоках, усл. ед.'
descriptive_statistics(
    full['rougher.input.feed_pb'], 'В сырье', col_name).join(
descriptive_statistics(
    full['rougher.output.concentrate_pb'], 'Со стадии флотации', col_name)).join(
descriptive_statistics(
    full['primary_cleaner.output.concentrate_pb'], 'Со стадии очистки 1', col_name)).join(
descriptive_statistics(
    full['final.output.concentrate_pb'], 'Во флотоконцентрате', col_name))


columns = ['rougher.input.feed_pb',
           'rougher.output.concentrate_pb',
           'primary_cleaner.output.concentrate_pb',
           'final.output.concentrate_pb']
label = ['в сырье',
         'со стадии флотации',
         'со стадии очистки 1',
         'во флотоконцентрате']


fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(10, 6))
ax1.set_title('Диаграмма распределения частот и диаграмма размаха' +
              '\nзначений содержания свинца в потоках')
for column in columns:
    ax1.hist(full[column], bins=100, alpha=0.5)
ax1.tick_params(labelbottom=False)
ax1.set_ylabel('Частота')
ax1.legend(label)
ax2.boxplot(full[columns], vert=False)
ax2.set_yticks(ticks=[1,2,3,4], labels=label)
ax2.set_xlabel('Содержание свинца, усл. ед.')
plt.show()


full['rougher.input.feed_me'] = full['rougher.input.feed_au'] + \
full['rougher.input.feed_ag'] + full['rougher.input.feed_pb']

full['rougher.output.concentrate_me'] = full['rougher.output.concentrate_au'] + \
full['rougher.output.concentrate_ag'] + full['rougher.output.concentrate_pb']

full['primary_cleaner.output.concentrate_me'] = \
full['primary_cleaner.output.concentrate_au'] + \
full['primary_cleaner.output.concentrate_ag'] + \
full['primary_cleaner.output.concentrate_pb']

full['final.output.concentrate_me'] = full['final.output.concentrate_au'] + \
full['final.output.concentrate_ag'] + full['final.output.concentrate_pb']


col_name = 'Содержание металлов в потоках, усл. ед.'
descriptive_statistics(
    full['rougher.input.feed_me'], 'В сырье', col_name).join(
descriptive_statistics(
    full['rougher.output.concentrate_me'], 'Со стадии флотации', col_name)).join(
descriptive_statistics(
    full['primary_cleaner.output.concentrate_me'], 'Со стадии очистки 1', col_name)).join(
descriptive_statistics(
    full['final.output.concentrate_me'], 'Во флотоконцентрате', col_name))


columns = ['rougher.input.feed_me',
           'rougher.output.concentrate_me',
           'primary_cleaner.output.concentrate_me',
           'final.output.concentrate_me']
label = ['в сырье',
         'со стадии флотации',
         'со стадии очистки 1',
         'во флотоконцентрате']


fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(10, 6))
ax1.set_title('Диаграмма распределения частот и диаграмма размаха' +
              '\nзначений содержания металлов в потоках')
for column in columns:
    ax1.hist(full[column], bins=100, alpha=0.5)
ax1.tick_params(labelbottom=False)
ax1.set_ylabel('Частота')
ax1.legend(label)
ax2.boxplot(full[columns], vert=False)
ax2.set_yticks(ticks=[1,2,3,4], labels=label)
ax2.set_xlabel('Содержание металлов, усл. ед.')
plt.show()


col_name = 'Размер гранул сырья, усл. ед.'
descriptive_statistics(
    test['rougher.input.feed_size'], 'В тестовой выборке', col_name).join(
descriptive_statistics(
    train['rougher.input.feed_size'], 'В тренировочной выборке', col_name))


label = ['В тестовой выборке', 'В тренировочной выборке']

fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(10, 6))
ax1.set_title('Диаграмма распределения частот и диаграмма размаха' +
              '\nзначений размера гранул сырья')
ax1.hist([test['rougher.input.feed_size'].tolist(),
          train['rougher.input.feed_size'].tolist()],
         bins=50, alpha=0.7)
ax1.legend(label)
ax1.tick_params(labelbottom=False)
ax1.set_ylabel('Частота')
ax1.legend(label)
ax2.boxplot([test['rougher.input.feed_size'].tolist(),
             train['rougher.input.feed_size'].tolist()],
            vert=False)
ax2.set_yticks(ticks=[1,2], labels=label)
ax2.set_xlabel('Размер гранул сырья, усл. ед.')
plt.show()


label = ['В тестовой выборке', 'В тренировочной выборке']

plt.figure(figsize=(10, 3))
plt.title('График ядерной оценки плотности распределения')
sns.kdeplot(test['rougher.input.feed_size'])
sns.kdeplot(train['rougher.input.feed_size'])
plt.xlabel('Размер гранул сырья, усл. ед.')
plt.ylabel('Плотность распределения')
plt.legend(label)
plt.ylim(-0.0005, 0.037)
plt.xlim(0, 495)
plt.show()


test.drop(test.loc[:, test.columns.str.contains('concentrate') |
                   test.columns.str.contains('tail') |
                   test.columns.str.contains('recovery')].columns,
          axis=1,
          inplace=True)
train = train.loc[:, test.columns.tolist()]


test.shape

(4626, 52)


train.shape

(12929, 52)


columns_model_rougher = test.loc[:, test.columns.str.contains('rougher')].columns
columns_model_rougher.tolist()

['rougher.input.feed_au',
 'rougher.input.feed_ag',
 'rougher.input.feed_pb',
 'rougher.input.feed_sol',
 'rougher.input.feed_rate',
 'rougher.input.feed_size',
 'rougher.input.floatbank10_sulfate',
 'rougher.input.floatbank10_xanthate',
 'rougher.state.floatbank10_a_air',
 'rougher.state.floatbank10_a_level',
 'rougher.state.floatbank10_b_air',
 'rougher.state.floatbank10_b_level',
 'rougher.state.floatbank10_c_air',
 'rougher.state.floatbank10_c_level',
 'rougher.state.floatbank10_d_air',
 'rougher.state.floatbank10_d_level',
 'rougher.state.floatbank10_e_air',
 'rougher.state.floatbank10_e_level',
 'rougher.state.floatbank10_f_air',
 'rougher.state.floatbank10_f_level',
 'rougher.input.floatbank11_sulfate',
 'rougher.input.floatbank11_xanthate']


features_rougher_train = train[columns_model_rougher]
target_rougher_train = full.loc[train.index.tolist(), 'rougher.output.recovery']

features_rougher_test = test[columns_model_rougher]
target_rougher_test = full.loc[test.index.tolist(), 'rougher.output.recovery']


features_rougher_train.shape

(12929, 22)


features_rougher_test.shape

(4626, 22)


features_final_train = train
target_final_train = full.loc[train.index, 'final.output.recovery']

features_final_test = test
target_final_test = full.loc[test.index, 'final.output.recovery']


features_final_train.shape

(12929, 52)


features_final_test.shape

(4626, 52)


def smape_score(target, predictions):
    '''
    Функция рассчитывает значение sMAPE.

    Принимает значения целевого признака и значения предсказания.
    Возвращает значение метрики sMAPE (в %).
    '''

    df = pd.DataFrame({'target': target.tolist(),
                       'predictions': predictions.tolist()})

    score = 2 * np.sum(
        abs(df['target'] - df['predictions']) /
        (abs(df['target']) + abs(df['predictions']))) / len(df) * 100

    return score


def scoring():

    return make_scorer(smape_score, greater_is_better=False)


def smape_final(rougher_target, final_target,
                rougher_predictions, final_predictions):
    '''
    Функция рассчитывает итоговое sMAPE для процесса флотационного обогащения.

    Принимает две обученные модели, два набора признаков для обучения
    и два целевых признака.
    Возвращает значение итоговой метрики (в %).
    '''

    score = 0.25 * smape_score(rougher_target, rougher_predictions) + \
            0.75 * smape_score(final_target, final_predictions)

    return round(score, 2)


round(train.shape[0]/test.shape[0])

3


CV = 3


%%time

MAX_TREE_DEPTH = 6

best_model_rougher_dtr = None
best_model_final_dtr = None
best_model_dtr_score = 100
best_tree_depth = 0

for depth in [None]+[i for i in range(1, MAX_TREE_DEPTH+1)]:
    model_rougher_dtr = DecisionTreeRegressor(max_depth=depth,
                                              random_state=RANDOM_STATE)
    model_final_dtr = DecisionTreeRegressor(max_depth=depth,
                                            random_state=RANDOM_STATE)

    model_dtr_score = 0.25 * cross_val_score(
                          model_rougher_dtr,
                          X=features_rougher_train,
                          y=target_rougher_train,
                          cv=CV,
                          scoring=scoring(),
                          n_jobs=-1).mean() + \
                      0.75 * cross_val_score(
                          model_final_dtr,
                          X=features_final_train,
                          y=target_final_train,
                          cv=CV,
                          scoring=scoring(),
                          n_jobs=-1).mean()

    if abs(model_dtr_score) < best_model_dtr_score:
        best_model_rougher_dtr = model_rougher_dtr
        best_model_final_dtr = model_rougher_dtr
        best_model_dtr_score = abs(model_dtr_score)
        best_tree_depth = depth

print('Модель решающего дерева')
print('Глубина дерева:', best_tree_depth)
print('sMAPE:', round(best_model_dtr_score, 2), '%')

Модель решающего дерева
Глубина дерева: 2
sMAPE: 8.6 %
CPU times: user 711 ms, sys: 130 ms, total: 841 ms
Wall time: 11.6 s


%%time

MAX_TREE_NUMBER = 100
MAX_TREE_DEPTH = 6

best_model_rougher_rfr = None
best_model_final_rfr = None
best_model_rfr_score = 100
best_tree_number = 0
best_tree_depth = 0

for est in range(10, MAX_TREE_NUMBER+1, 10):
    for depth in [None]+[i for i in range(1, MAX_TREE_DEPTH+1)]:
        model_rougher_rfr = RandomForestRegressor(n_estimators=est,
                                                  max_depth=depth,
                                                  random_state=RANDOM_STATE,
                                                  n_jobs=-1)
        model_final_rfr = RandomForestRegressor(n_estimators=est,
                                                max_depth=depth,
                                                random_state=RANDOM_STATE,
                                                n_jobs=-1)

        model_rfr_score = 0.25 * cross_val_score(
                              model_rougher_rfr,
                              X=features_rougher_train,
                              y=target_rougher_train,
                              cv=CV,
                              scoring=scoring(),
                              n_jobs=-1).mean() + \
                          0.75 * cross_val_score(
                              model_final_rfr,
                              X=features_final_train,
                              y=target_final_train,
                              cv=CV,
                              scoring=scoring(),
                              n_jobs=-1).mean()

        if abs(model_rfr_score) < best_model_rfr_score:
            best_model_rougher_rfr = model_rougher_rfr
            best_model_final_rfr = model_final_rfr
            best_model_rfr_score = abs(model_rfr_score)
            best_tree_number = est
            best_tree_depth = depth

print('Модель случайного леса')
print('Количество деревьев в лесу:', best_tree_number)
print('Глубина каждого дерева:', best_tree_depth)
print('sMAPE:', round(best_model_rfr_score, 2), '%')

Модель случайного леса
Количество деревьев в лесу: 80
Глубина каждого дерева: 2
sMAPE: 8.49 %
CPU times: user 26 s, sys: 4.12 s, total: 30.1 s
Wall time: 40min 48s


%%time

model_rougher_lr = LinearRegression(n_jobs=-1)
model_final_lr = LinearRegression(n_jobs=-1)

model_lr_score = 0.25 * cross_val_score(
                    model_rougher_lr,
                    X=features_rougher_train,
                    y=target_rougher_train,
                    cv=CV,
                    scoring=scoring(),
                    n_jobs=-1).mean() + \
                 0.75 * cross_val_score(
                    model_final_lr,
                    X=features_final_train,
                    y=target_final_train,
                    cv=CV,
                    scoring=scoring(),
                    n_jobs=-1).mean()

print('Модель линейной регрессии')
print('sMAPE', round(abs(model_lr_score), 2), '%')

Модель линейной регрессии
sMAPE 9.88 %
CPU times: user 95.3 ms, sys: 8.55 ms, total: 104 ms
Wall time: 466 ms


best_model_rougher_rfr.fit(features_rougher_train, target_rougher_train)
best_model_final_rfr.fit(features_final_train, target_final_train)

rougher_predictions_rfr = best_model_rougher_rfr.predict(features_rougher_test)
final_predictions_rfr = best_model_final_rfr.predict(features_final_test)

smape = smape_final(target_rougher_test, target_final_test,
                    rougher_predictions_rfr, final_predictions_rfr)

print('Модель случайного леса')
print('Количество деревьев в лесу: 80')
print('Глубина каждого дерева: 2')
print('sMAPE:', smape, '%')

Модель случайного леса
Количество деревьев в лесу: 80
Глубина каждого дерева: 2
sMAPE: 7.5 %


STRATEGY = 'mean'

model_rougher_dr = DummyRegressor(strategy=STRATEGY)
model_final_dr = DummyRegressor(strategy=STRATEGY)

model_rougher_dr.fit(X=features_rougher_train, y=target_rougher_train)
model_final_dr.fit(X=features_final_train, y=target_final_train)

target_rougher_predicted_dr = model_rougher_dr.predict(X=features_rougher_test)
target_final_predicted_dr = model_final_dr.predict(X=features_final_test)

smape = smape_final(target_rougher_test, target_final_test,
                    target_rougher_predicted_dr, target_final_predicted_dr)

print('Простая модель, предсказывающая среднее арифметическое значение')
print('sMAPE:', smape, '%')

Простая модель, предсказывающая среднее арифметическое значение
sMAPE: 8.35 %

	date	rougher.input.feed_au	rougher.input.feed_ag	rougher.input.feed_pb	rougher.input.feed_sol	rougher.input.feed_rate	rougher.input.feed_size	rougher.input.floatbank10_sulfate	rougher.input.floatbank10_xanthate	rougher.state.floatbank10_a_air	rougher.state.floatbank10_a_level	rougher.state.floatbank10_b_air	rougher.state.floatbank10_b_level	rougher.state.floatbank10_c_air	rougher.state.floatbank10_c_level	rougher.state.floatbank10_d_air	rougher.state.floatbank10_d_level	rougher.state.floatbank10_e_air	rougher.state.floatbank10_e_level	rougher.state.floatbank10_f_air	rougher.state.floatbank10_f_level	rougher.input.floatbank11_sulfate	rougher.input.floatbank11_xanthate	rougher.calculation.sulfate_to_au_concentrate	rougher.calculation.floatbank10_sulfate_to_au_feed	rougher.calculation.floatbank11_sulfate_to_au_feed	rougher.calculation.au_pb_ratio	rougher.output.concentrate_au	rougher.output.concentrate_ag	rougher.output.concentrate_pb	rougher.output.concentrate_sol	rougher.output.recovery	rougher.output.tail_au	rougher.output.tail_ag	rougher.output.tail_pb	rougher.output.tail_sol	primary_cleaner.input.sulfate	primary_cleaner.input.depressant	primary_cleaner.input.feed_size	primary_cleaner.input.xanthate	primary_cleaner.state.floatbank8_a_air	primary_cleaner.state.floatbank8_a_level	primary_cleaner.state.floatbank8_b_air	primary_cleaner.state.floatbank8_b_level	primary_cleaner.state.floatbank8_c_air	primary_cleaner.state.floatbank8_c_level	primary_cleaner.state.floatbank8_d_air	primary_cleaner.state.floatbank8_d_level	primary_cleaner.output.concentrate_au	primary_cleaner.output.concentrate_ag	primary_cleaner.output.concentrate_pb	primary_cleaner.output.concentrate_sol	primary_cleaner.output.tail_au	primary_cleaner.output.tail_ag	primary_cleaner.output.tail_pb	primary_cleaner.output.tail_sol	secondary_cleaner.state.floatbank2_a_air	secondary_cleaner.state.floatbank2_a_level	secondary_cleaner.state.floatbank2_b_air	secondary_cleaner.state.floatbank2_b_level	secondary_cleaner.state.floatbank3_a_air	secondary_cleaner.state.floatbank3_a_level	secondary_cleaner.state.floatbank3_b_air	secondary_cleaner.state.floatbank3_b_level	secondary_cleaner.state.floatbank4_a_air	secondary_cleaner.state.floatbank4_a_level	secondary_cleaner.state.floatbank4_b_air	secondary_cleaner.state.floatbank4_b_level	secondary_cleaner.state.floatbank5_a_air	secondary_cleaner.state.floatbank5_a_level	secondary_cleaner.state.floatbank5_b_air	secondary_cleaner.state.floatbank5_b_level	secondary_cleaner.state.floatbank6_a_air	secondary_cleaner.state.floatbank6_a_level	secondary_cleaner.output.tail_au	secondary_cleaner.output.tail_ag	secondary_cleaner.output.tail_pb	secondary_cleaner.output.tail_sol	final.output.concentrate_au	final.output.concentrate_ag	final.output.concentrate_pb	final.output.concentrate_sol	final.output.recovery	final.output.tail_au	final.output.tail_ag	final.output.tail_pb	final.output.tail_sol
0	2016-01-15 00:00:00	6.486150	6.100378	2.284912	36.808594	523.546326	55.486599	11.986616	6.007990	999.706909	-404.066986	1603.011353	-434.715027	1602.37500	-442.204468	1598.937256	-451.294128	1404.472046	-455.462982	1416.354980	-451.939636	11.836743	6.005818	41885.707031	3481.779053	3520.337158	2.838687	19.793808	11.500771	7.101074	28.029297	87.107763	1.170244	5.008018	0.508728	19.154297	127.092003	10.128295	7.25	0.988759	1549.775757	-498.912140	1551.434204	-516.403442	1549.873901	-498.666595	1554.367432	-493.428131	34.174427	8.547551	10.389648	19.529297	2.106679	14.936526	2.534912	7.476074	25.853109	-498.526489	23.89366	-501.406281	23.961798	-495.262817	21.940409	-499.340973	14.016835	-502.488007	12.099931	-504.715942	9.925633	-498.310211	8.079666	-500.470978	14.151341	-605.841980	2.606185	14.500184	4.694824	8.764648	42.192020	6.055403	9.889648	5.507324	70.541216	2.143149	10.411962	0.895447	16.904297
1	2016-01-15 01:00:00	6.478583	6.161113	2.266033	35.753385	525.290581	57.278666	11.971193	6.005766	1000.286398	-400.065196	1600.754587	-449.953435	1600.47958	-449.830646	1600.527589	-449.953649	1399.227084	-450.869848	1399.719514	-450.119001	11.996163	6.012594	42050.861809	3498.370979	3489.981857	2.858998	20.050975	11.615865	7.278807	28.067063	86.843261	1.184827	4.955378	0.536689	18.965234	125.629232	10.296251	7.25	1.002663	1576.166671	-500.904965	1575.950626	-499.865889	1575.994189	-499.315107	1574.479259	-498.931665	34.118526	8.558743	10.497069	19.369147	2.353017	16.250511	3.049646	6.733907	25.880539	-499.989656	23.88953	-500.372428	23.970550	-500.085473	22.085714	-499.446897	13.992281	-505.503262	11.950531	-501.331529	10.039245	-500.169983	7.984757	-500.582168	13.998353	-599.787184	2.488248	14.265474	4.592547	9.001494	42.701629	6.029369	9.968944	5.257781	69.266198	2.224930	10.462676	0.927452	16.634514

	date	rougher.input.feed_au	rougher.input.feed_ag	rougher.input.feed_pb	rougher.input.feed_sol	rougher.input.feed_rate	rougher.input.feed_size	rougher.input.floatbank10_sulfate	rougher.input.floatbank10_xanthate	rougher.state.floatbank10_a_air	rougher.state.floatbank10_a_level	rougher.state.floatbank10_b_air	rougher.state.floatbank10_b_level	rougher.state.floatbank10_c_air	rougher.state.floatbank10_c_level	rougher.state.floatbank10_d_air	rougher.state.floatbank10_d_level	rougher.state.floatbank10_e_air	rougher.state.floatbank10_e_level	rougher.state.floatbank10_f_air	rougher.state.floatbank10_f_level	rougher.input.floatbank11_sulfate	rougher.input.floatbank11_xanthate	rougher.calculation.sulfate_to_au_concentrate	rougher.calculation.floatbank10_sulfate_to_au_feed	rougher.calculation.floatbank11_sulfate_to_au_feed	rougher.calculation.au_pb_ratio	rougher.output.concentrate_au	rougher.output.concentrate_ag	rougher.output.concentrate_pb	rougher.output.concentrate_sol	rougher.output.recovery	rougher.output.tail_au	rougher.output.tail_ag	rougher.output.tail_pb	rougher.output.tail_sol	primary_cleaner.input.sulfate	primary_cleaner.input.depressant	primary_cleaner.input.feed_size	primary_cleaner.input.xanthate	primary_cleaner.state.floatbank8_a_air	primary_cleaner.state.floatbank8_a_level	primary_cleaner.state.floatbank8_b_air	primary_cleaner.state.floatbank8_b_level	primary_cleaner.state.floatbank8_c_air	primary_cleaner.state.floatbank8_c_level	primary_cleaner.state.floatbank8_d_air	primary_cleaner.state.floatbank8_d_level	primary_cleaner.output.concentrate_au	primary_cleaner.output.concentrate_ag	primary_cleaner.output.concentrate_pb	primary_cleaner.output.concentrate_sol	primary_cleaner.output.tail_au	primary_cleaner.output.tail_ag	primary_cleaner.output.tail_pb	primary_cleaner.output.tail_sol	secondary_cleaner.state.floatbank2_a_air	secondary_cleaner.state.floatbank2_a_level	secondary_cleaner.state.floatbank2_b_air	secondary_cleaner.state.floatbank2_b_level	secondary_cleaner.state.floatbank3_a_air	secondary_cleaner.state.floatbank3_a_level	secondary_cleaner.state.floatbank3_b_air	secondary_cleaner.state.floatbank3_b_level	secondary_cleaner.state.floatbank4_a_air	secondary_cleaner.state.floatbank4_a_level	secondary_cleaner.state.floatbank4_b_air	secondary_cleaner.state.floatbank4_b_level	secondary_cleaner.state.floatbank5_a_air	secondary_cleaner.state.floatbank5_a_level	secondary_cleaner.state.floatbank5_b_air	secondary_cleaner.state.floatbank5_b_level	secondary_cleaner.state.floatbank6_a_air	secondary_cleaner.state.floatbank6_a_level	secondary_cleaner.output.tail_au	secondary_cleaner.output.tail_ag	secondary_cleaner.output.tail_pb	secondary_cleaner.output.tail_sol	final.output.concentrate_au	final.output.concentrate_ag	final.output.concentrate_pb	final.output.concentrate_sol	final.output.recovery	final.output.tail_au	final.output.tail_ag	final.output.tail_pb	final.output.tail_sol
14577	2017-12-09 12:59:59	4.838619	6.121323	4.144989	38.591551	559.031805	119.499241	7.356687	9.304952	1196.569267	-299.512227	1147.675196	-500.608341	1048.565741	-500.932810	949.773589	-500.023144	848.515225	-500.289405	850.013123	-496.822119	7.095508	9.297924	46200.099585	2614.403937	2518.550078	1.167342	18.441436	11.425983	10.888213	29.783972	87.724007	0.771582	4.177943	0.650557	16.960649	120.878188	7.962636	6.5	1.302419	1649.820162	-399.930973	1649.357538	-399.721222	1648.656192	-401.195834	1649.725133	-400.636306	32.925325	6.612376	12.623947	7.542264	2.462890	23.002490	6.629725	4.089124	35.026062	-499.891945	29.921795	-499.949663	26.031747	-500.384612	22.991058	-500.079590	22.960095	-501.612783	20.035660	-500.251357	17.998535	-500.395178	12.954048	-499.895163	19.968498	-501.041608	0.0	0.0	0.0	0.0	46.866780	3.195978	11.349355	6.862249	69.049291	1.612542	8.897321	3.130493	10.549470
14578	2017-12-09 13:59:59	4.525061	5.970515	4.020002	40.074026	555.682872	122.262690	6.586020	9.299606	1204.866639	-299.235675	1149.942902	-501.717903	1049.604390	-500.549053	952.702732	-502.352296	849.016017	-500.505677	850.455635	-506.897968	6.584130	9.300133	44585.181334	2510.813545	2510.142433	1.125637	15.111231	8.523497	8.955135	22.786467	88.890579	0.685059	4.050010	0.636095	18.323080	105.666118	7.955111	6.5	1.315926	1649.166761	-399.888631	1649.196904	-399.677571	1647.896999	-399.988275	1649.772714	-399.831902	31.856742	7.128896	12.633558	7.941900	2.612231	22.298149	6.278250	4.457457	35.003586	-501.083794	29.990533	-611.855898	25.948429	-500.067268	22.968268	-499.839442	23.015718	-501.711599	19.951231	-499.857027	18.019543	-500.451156	13.023431	-499.914391	19.990885	-501.518452	0.0	0.0	0.0	0.0	46.795691	3.109998	11.434366	6.886013	67.002189	1.596616	8.529606	2.911418	11.115147

	date	rougher.input.feed_au	rougher.input.feed_ag	rougher.input.feed_pb	rougher.input.feed_sol	rougher.input.feed_rate	rougher.input.feed_size	rougher.input.floatbank10_sulfate	rougher.input.floatbank10_xanthate	rougher.state.floatbank10_a_air	rougher.state.floatbank10_a_level	rougher.state.floatbank10_b_air	rougher.state.floatbank10_b_level	rougher.state.floatbank10_c_air	rougher.state.floatbank10_c_level	rougher.state.floatbank10_d_air	rougher.state.floatbank10_d_level	rougher.state.floatbank10_e_air	rougher.state.floatbank10_e_level	rougher.state.floatbank10_f_air	rougher.state.floatbank10_f_level	rougher.input.floatbank11_sulfate	rougher.input.floatbank11_xanthate	primary_cleaner.input.sulfate	primary_cleaner.input.depressant	primary_cleaner.input.feed_size	primary_cleaner.input.xanthate	primary_cleaner.state.floatbank8_a_air	primary_cleaner.state.floatbank8_a_level	primary_cleaner.state.floatbank8_b_air	primary_cleaner.state.floatbank8_b_level	primary_cleaner.state.floatbank8_c_air	primary_cleaner.state.floatbank8_c_level	primary_cleaner.state.floatbank8_d_air	primary_cleaner.state.floatbank8_d_level	secondary_cleaner.state.floatbank2_a_air	secondary_cleaner.state.floatbank2_a_level	secondary_cleaner.state.floatbank2_b_air	secondary_cleaner.state.floatbank2_b_level	secondary_cleaner.state.floatbank3_a_air	secondary_cleaner.state.floatbank3_a_level	secondary_cleaner.state.floatbank3_b_air	secondary_cleaner.state.floatbank3_b_level	secondary_cleaner.state.floatbank4_a_air	secondary_cleaner.state.floatbank4_a_level	secondary_cleaner.state.floatbank4_b_air	secondary_cleaner.state.floatbank4_b_level	secondary_cleaner.state.floatbank5_a_air	secondary_cleaner.state.floatbank5_a_level	secondary_cleaner.state.floatbank5_b_air	secondary_cleaner.state.floatbank5_b_level	secondary_cleaner.state.floatbank6_a_air	secondary_cleaner.state.floatbank6_a_level
0	2017-12-09 14:59:59	4.365491	6.158718	3.875727	39.135119	555.820208	94.544358	6.146982	9.308612	1196.238112	-299.862743	1147.248241	-500.363165	1047.279065	-500.354091	948.756608	-498.439416	849.441918	-499.255503	850.112246	-499.407112	6.168939	9.309852	95.248427	8.078957	6.5	1.283045	1648.759906	-399.862053	1650.135395	-399.957321	1648.831890	-400.586116	1649.464582	-400.673303	34.940919	-500.150510	30.031867	-500.328335	26.033990	-500.147792	22.952306	-500.037678	23.018622	-500.492702	20.020205	-500.220296	17.963512	-499.939490	12.990306	-500.080993	19.990336	-499.191575
1	2017-12-09 15:59:59	4.362781	6.048130	3.902537	39.713906	544.731687	123.742430	6.210119	9.297709	1201.904177	-299.907308	1154.087804	-500.036580	1054.009756	-500.237335	944.138793	-496.866953	851.589767	-499.040466	851.345606	-499.122561	6.209517	9.297194	98.880538	7.984164	6.5	1.241969	1646.547763	-398.977083	1648.212240	-400.383265	1648.917387	-399.521344	1651.498591	-399.745329	34.980742	-498.131002	29.968453	-586.013330	25.971737	-499.608392	22.958448	-499.821308	23.024963	-501.153409	20.054122	-500.314711	17.979515	-499.272871	12.992404	-499.976268	20.013986	-500.625471

	date	rougher.input.feed_au	rougher.input.feed_ag	rougher.input.feed_pb	rougher.input.feed_sol	rougher.input.feed_rate	rougher.input.feed_size	rougher.input.floatbank10_sulfate	rougher.input.floatbank10_xanthate	rougher.state.floatbank10_a_air	rougher.state.floatbank10_a_level	rougher.state.floatbank10_b_air	rougher.state.floatbank10_b_level	rougher.state.floatbank10_c_air	rougher.state.floatbank10_c_level	rougher.state.floatbank10_d_air	rougher.state.floatbank10_d_level	rougher.state.floatbank10_e_air	rougher.state.floatbank10_e_level	rougher.state.floatbank10_f_air	rougher.state.floatbank10_f_level	rougher.input.floatbank11_sulfate	rougher.input.floatbank11_xanthate	primary_cleaner.input.sulfate	primary_cleaner.input.depressant	primary_cleaner.input.feed_size	primary_cleaner.input.xanthate	primary_cleaner.state.floatbank8_a_air	primary_cleaner.state.floatbank8_a_level	primary_cleaner.state.floatbank8_b_air	primary_cleaner.state.floatbank8_b_level	primary_cleaner.state.floatbank8_c_air	primary_cleaner.state.floatbank8_c_level	primary_cleaner.state.floatbank8_d_air	primary_cleaner.state.floatbank8_d_level	secondary_cleaner.state.floatbank2_a_air	secondary_cleaner.state.floatbank2_a_level	secondary_cleaner.state.floatbank2_b_air	secondary_cleaner.state.floatbank2_b_level	secondary_cleaner.state.floatbank3_a_air	secondary_cleaner.state.floatbank3_a_level	secondary_cleaner.state.floatbank3_b_air	secondary_cleaner.state.floatbank3_b_level	secondary_cleaner.state.floatbank4_a_air	secondary_cleaner.state.floatbank4_a_level	secondary_cleaner.state.floatbank4_b_air	secondary_cleaner.state.floatbank4_b_level	secondary_cleaner.state.floatbank5_a_air	secondary_cleaner.state.floatbank5_a_level	secondary_cleaner.state.floatbank5_b_air	secondary_cleaner.state.floatbank5_b_level	secondary_cleaner.state.floatbank6_a_air	secondary_cleaner.state.floatbank6_a_level
4858	2018-08-18 09:59:59	6.702243	6.097592	2.278955	37.728375	549.812490	59.637037	13.746510	6.706417	999.635748	-399.484884	1599.852212	-449.448826	1599.325946	-448.812166	1614.612892	-449.774665	1401.626496	-447.607440	1400.158685	-448.035466	13.758277	6.702969	147.944885	12.481529	7.25	0.893913	1603.053015	-499.728573	1600.537532	-500.096021	1597.568005	-499.601040	1599.213513	-500.964698	26.111929	-499.496608	24.094419	-500.250900	23.945513	-500.816108	23.995434	-499.761541	13.983963	-503.675714	11.928745	-500.583183	9.970359	-498.822343	8.016377	-499.013145	14.000032	-600.024741
4859	2018-08-18 10:59:59	6.633002	5.908793	2.440899	36.958540	573.076049	58.509262	13.707775	6.820245	1000.589156	-398.879586	1598.975250	-451.238638	1600.058139	-451.271448	1610.182874	-450.460190	1399.194663	-450.332681	1399.638290	-450.198722	13.812674	6.790624	138.120409	12.464032	7.25	0.898779	1603.436598	-499.864400	1597.455353	-500.266905	1600.747882	-500.541002	1599.754053	-499.216962	26.059510	-499.151646	23.934570	-499.351758	23.994184	-500.078566	23.924481	-499.989024	13.989632	-503.195299	12.052233	-500.928547	9.962574	-502.986357	7.979219	-500.146835	13.981614	-598.070855

	date	rougher.input.feed_au	rougher.input.feed_ag	rougher.input.feed_pb	rougher.input.feed_sol	rougher.input.feed_rate	rougher.input.feed_size	rougher.input.floatbank10_sulfate	rougher.input.floatbank10_xanthate	rougher.state.floatbank10_a_air	rougher.state.floatbank10_a_level	rougher.state.floatbank10_b_air	rougher.state.floatbank10_b_level	rougher.state.floatbank10_c_air	rougher.state.floatbank10_c_level	rougher.state.floatbank10_d_air	rougher.state.floatbank10_d_level	rougher.state.floatbank10_e_air	rougher.state.floatbank10_e_level	rougher.state.floatbank10_f_air	rougher.state.floatbank10_f_level	rougher.input.floatbank11_sulfate	rougher.input.floatbank11_xanthate	rougher.calculation.sulfate_to_au_concentrate	rougher.calculation.floatbank10_sulfate_to_au_feed	rougher.calculation.floatbank11_sulfate_to_au_feed	rougher.calculation.au_pb_ratio	rougher.output.concentrate_au	rougher.output.concentrate_ag	rougher.output.concentrate_pb	rougher.output.concentrate_sol	rougher.output.recovery	rougher.output.tail_au	rougher.output.tail_ag	rougher.output.tail_pb	rougher.output.tail_sol	primary_cleaner.input.sulfate	primary_cleaner.input.depressant	primary_cleaner.input.feed_size	primary_cleaner.input.xanthate	primary_cleaner.state.floatbank8_a_air	primary_cleaner.state.floatbank8_a_level	primary_cleaner.state.floatbank8_b_air	primary_cleaner.state.floatbank8_b_level	primary_cleaner.state.floatbank8_c_air	primary_cleaner.state.floatbank8_c_level	primary_cleaner.state.floatbank8_d_air	primary_cleaner.state.floatbank8_d_level	primary_cleaner.output.concentrate_au	primary_cleaner.output.concentrate_ag	primary_cleaner.output.concentrate_pb	primary_cleaner.output.concentrate_sol	primary_cleaner.output.tail_au	primary_cleaner.output.tail_ag	primary_cleaner.output.tail_pb	primary_cleaner.output.tail_sol	secondary_cleaner.state.floatbank2_a_air	secondary_cleaner.state.floatbank2_a_level	secondary_cleaner.state.floatbank2_b_air	secondary_cleaner.state.floatbank2_b_level	secondary_cleaner.state.floatbank3_a_air	secondary_cleaner.state.floatbank3_a_level	secondary_cleaner.state.floatbank3_b_air	secondary_cleaner.state.floatbank3_b_level	secondary_cleaner.state.floatbank4_a_air	secondary_cleaner.state.floatbank4_a_level	secondary_cleaner.state.floatbank4_b_air	secondary_cleaner.state.floatbank4_b_level	secondary_cleaner.state.floatbank5_a_air	secondary_cleaner.state.floatbank5_a_level	secondary_cleaner.state.floatbank5_b_air	secondary_cleaner.state.floatbank5_b_level	secondary_cleaner.state.floatbank6_a_air	secondary_cleaner.state.floatbank6_a_level	secondary_cleaner.output.tail_au	secondary_cleaner.output.tail_ag	secondary_cleaner.output.tail_pb	secondary_cleaner.output.tail_sol	final.output.concentrate_au	final.output.concentrate_ag	final.output.concentrate_pb	final.output.concentrate_sol	final.output.recovery	final.output.tail_au	final.output.tail_ag	final.output.tail_pb	final.output.tail_sol
0	2016-01-15 00:00:00	6.486150	6.100378	2.284912	36.808594	523.546326	55.486599	11.986616	6.007990	999.706909	-404.066986	1603.011353	-434.715027	1602.37500	-442.204468	1598.937256	-451.294128	1404.472046	-455.462982	1416.354980	-451.939636	11.836743	6.005818	41885.707031	3481.779053	3520.337158	2.838687	19.793808	11.500771	7.101074	28.029297	87.107763	1.170244	5.008018	0.508728	19.154297	127.092003	10.128295	7.25	0.988759	1549.775757	-498.912140	1551.434204	-516.403442	1549.873901	-498.666595	1554.367432	-493.428131	34.174427	8.547551	10.389648	19.529297	2.106679	14.936526	2.534912	7.476074	25.853109	-498.526489	23.89366	-501.406281	23.961798	-495.262817	21.940409	-499.340973	14.016835	-502.488007	12.099931	-504.715942	9.925633	-498.310211	8.079666	-500.470978	14.151341	-605.841980	2.606185	14.500184	4.694824	8.764648	42.192020	6.055403	9.889648	5.507324	70.541216	2.143149	10.411962	0.895447	16.904297
1	2016-01-15 01:00:00	6.478583	6.161113	2.266033	35.753385	525.290581	57.278666	11.971193	6.005766	1000.286398	-400.065196	1600.754587	-449.953435	1600.47958	-449.830646	1600.527589	-449.953649	1399.227084	-450.869848	1399.719514	-450.119001	11.996163	6.012594	42050.861809	3498.370979	3489.981857	2.858998	20.050975	11.615865	7.278807	28.067063	86.843261	1.184827	4.955378	0.536689	18.965234	125.629232	10.296251	7.25	1.002663	1576.166671	-500.904965	1575.950626	-499.865889	1575.994189	-499.315107	1574.479259	-498.931665	34.118526	8.558743	10.497069	19.369147	2.353017	16.250511	3.049646	6.733907	25.880539	-499.989656	23.88953	-500.372428	23.970550	-500.085473	22.085714	-499.446897	13.992281	-505.503262	11.950531	-501.331529	10.039245	-500.169983	7.984757	-500.582168	13.998353	-599.787184	2.488248	14.265474	4.592547	9.001494	42.701629	6.029369	9.968944	5.257781	69.266198	2.224930	10.462676	0.927452	16.634514

	date	rougher.input.feed_au	rougher.input.feed_ag	rougher.input.feed_pb	rougher.input.feed_sol	rougher.input.feed_rate	rougher.input.feed_size	rougher.input.floatbank10_sulfate	rougher.input.floatbank10_xanthate	rougher.state.floatbank10_a_air	rougher.state.floatbank10_a_level	rougher.state.floatbank10_b_air	rougher.state.floatbank10_b_level	rougher.state.floatbank10_c_air	rougher.state.floatbank10_c_level	rougher.state.floatbank10_d_air	rougher.state.floatbank10_d_level	rougher.state.floatbank10_e_air	rougher.state.floatbank10_e_level	rougher.state.floatbank10_f_air	rougher.state.floatbank10_f_level	rougher.input.floatbank11_sulfate	rougher.input.floatbank11_xanthate	rougher.calculation.sulfate_to_au_concentrate	rougher.calculation.floatbank10_sulfate_to_au_feed	rougher.calculation.floatbank11_sulfate_to_au_feed	rougher.calculation.au_pb_ratio	rougher.output.concentrate_au	rougher.output.concentrate_ag	rougher.output.concentrate_pb	rougher.output.concentrate_sol	rougher.output.recovery	rougher.output.tail_au	rougher.output.tail_ag	rougher.output.tail_pb	rougher.output.tail_sol	primary_cleaner.input.sulfate	primary_cleaner.input.depressant	primary_cleaner.input.feed_size	primary_cleaner.input.xanthate	primary_cleaner.state.floatbank8_a_air	primary_cleaner.state.floatbank8_a_level	primary_cleaner.state.floatbank8_b_air	primary_cleaner.state.floatbank8_b_level	primary_cleaner.state.floatbank8_c_air	primary_cleaner.state.floatbank8_c_level	primary_cleaner.state.floatbank8_d_air	primary_cleaner.state.floatbank8_d_level	primary_cleaner.output.concentrate_au	primary_cleaner.output.concentrate_ag	primary_cleaner.output.concentrate_pb	primary_cleaner.output.concentrate_sol	primary_cleaner.output.tail_au	primary_cleaner.output.tail_ag	primary_cleaner.output.tail_pb	primary_cleaner.output.tail_sol	secondary_cleaner.state.floatbank2_a_air	secondary_cleaner.state.floatbank2_a_level	secondary_cleaner.state.floatbank2_b_air	secondary_cleaner.state.floatbank2_b_level	secondary_cleaner.state.floatbank3_a_air	secondary_cleaner.state.floatbank3_a_level	secondary_cleaner.state.floatbank3_b_air	secondary_cleaner.state.floatbank3_b_level	secondary_cleaner.state.floatbank4_a_air	secondary_cleaner.state.floatbank4_a_level	secondary_cleaner.state.floatbank4_b_air	secondary_cleaner.state.floatbank4_b_level	secondary_cleaner.state.floatbank5_a_air	secondary_cleaner.state.floatbank5_a_level	secondary_cleaner.state.floatbank5_b_air	secondary_cleaner.state.floatbank5_b_level	secondary_cleaner.state.floatbank6_a_air	secondary_cleaner.state.floatbank6_a_level	secondary_cleaner.output.tail_au	secondary_cleaner.output.tail_ag	secondary_cleaner.output.tail_pb	secondary_cleaner.output.tail_sol	final.output.concentrate_au	final.output.concentrate_ag	final.output.concentrate_pb	final.output.concentrate_sol	final.output.recovery	final.output.tail_au	final.output.tail_ag	final.output.tail_pb	final.output.tail_sol
19437	2018-08-18 09:59:59	6.702243	6.097592	2.278955	37.728375	549.812490	59.637037	13.746510	6.706417	999.635748	-399.484884	1599.852212	-449.448826	1599.325946	-448.812166	1614.612892	-449.774665	1401.626496	-447.607440	1400.158685	-448.035466	13.758277	6.702969	42019.291707	3502.637709	3495.776643	2.940927	19.620480	11.528770	7.091962	27.774855	89.299981	1.031923	4.559246	0.557099	20.067665	147.944885	12.481529	7.25	0.893913	1603.053015	-499.728573	1600.537532	-500.096021	1597.568005	-499.601040	1599.213513	-500.964698	28.081869	5.899284	8.154167	14.301345	1.874228	15.993499	3.077749	7.312664	26.111929	-499.496608	24.094419	-500.250900	23.945513	-500.816108	23.995434	-499.761541	13.983963	-503.675714	11.928745	-500.583183	9.970359	-498.822343	8.016377	-499.013145	14.000032	-600.024741	1.868321	13.795229	4.576720	9.876374	43.339714	5.664983	10.168953	5.484318	72.659046	2.064422	10.475957	1.050592	17.525656
19438	2018-08-18 10:59:59	6.633002	5.908793	2.440899	36.958540	573.076049	58.509262	13.707775	6.820245	1000.589156	-398.879586	1598.975250	-451.238638	1600.058139	-451.271448	1610.182874	-450.460190	1399.194663	-450.332681	1399.638290	-450.198722	13.812674	6.790624	41882.126352	3499.915439	3519.952954	2.717442	19.497669	11.116297	7.297363	27.497520	87.650868	1.167140	4.667157	0.615418	18.907179	138.120409	12.464032	7.25	0.898779	1603.436598	-499.864400	1597.455353	-500.266905	1600.747882	-500.541002	1599.754053	-499.216962	34.826500	8.098985	10.525324	18.436589	2.025454	15.535214	2.866688	7.265906	26.059510	-499.151646	23.934570	-499.351758	23.994184	-500.078566	23.924481	-499.989024	13.989632	-503.195299	12.052233	-500.928547	9.962574	-502.986357	7.979219	-500.146835	13.981614	-598.070855	2.073737	13.209191	4.550553	9.813732	43.235220	5.687063	9.980404	5.282514	70.279619	2.209593	10.487013	1.159805	17.483979

	Всего пропусков	%
secondary_cleaner.output.tail_sol	1748	9.0
rougher.input.floatbank11_xanthate	453	2.3
rougher.state.floatbank10_e_air	436	2.2
primary_cleaner.output.concentrate_sol	370	1.9
secondary_cleaner.state.floatbank2_a_air	220	1.1
final.output.concentrate_sol	211	1.1
rougher.input.feed_size	145	0.7
primary_cleaner.output.concentrate_pb	116	0.6
primary_cleaner.input.xanthate	104	0.5
final.output.tail_pb	101	0.5
rougher.input.feed_pb	100	0.5
rougher.input.feed_sol	99	0.5
primary_cleaner.output.tail_sol	62	0.3
rougher.input.floatbank11_sulfate	44	0.2
primary_cleaner.input.depressant	37	0.2
rougher.input.floatbank10_sulfate	34	0.2
primary_cleaner.input.sulfate	24	0.1
rougher.output.concentrate_sol	23	0.1
secondary_cleaner.state.floatbank2_b_air	23	0.1
primary_cleaner.output.tail_pb	21	0.1
secondary_cleaner.state.floatbank3_a_air	13	0.1
secondary_cleaner.output.tail_pb	12	0.1
rougher.input.feed_rate	11	0.1
rougher.input.floatbank10_xanthate	8	0.0
secondary_cleaner.state.floatbank4_a_air	6	0.0
final.output.tail_sol	6	0.0
primary_cleaner.output.tail_ag	4	0.0
primary_cleaner.state.floatbank8_b_air	4	0.0
primary_cleaner.state.floatbank8_a_air	4	0.0
primary_cleaner.state.floatbank8_d_air	3	0.0
rougher.calculation.sulfate_to_au_concentrate	2	0.0
secondary_cleaner.output.tail_ag	2	0.0
primary_cleaner.state.floatbank8_c_air	2	0.0
secondary_cleaner.state.floatbank6_a_air	2	0.0
rougher.calculation.floatbank11_sulfate_to_au_feed	2	0.0
rougher.calculation.floatbank10_sulfate_to_au_feed	2	0.0
secondary_cleaner.state.floatbank3_b_air	1	0.0
final.output.tail_ag	1	0.0
final.output.concentrate_pb	1	0.0
final.output.concentrate_ag	1	0.0
secondary_cleaner.state.floatbank4_b_level	1	0.0
secondary_cleaner.state.floatbank4_b_air	1	0.0
secondary_cleaner.state.floatbank2_a_level	1	0.0
secondary_cleaner.state.floatbank6_a_level	1	0.0
secondary_cleaner.state.floatbank3_a_level	1	0.0
secondary_cleaner.state.floatbank4_a_level	1	0.0
secondary_cleaner.state.floatbank2_b_level	1	0.0
secondary_cleaner.state.floatbank5_a_air	1	0.0
secondary_cleaner.state.floatbank5_b_level	1	0.0
secondary_cleaner.state.floatbank5_b_air	1	0.0
secondary_cleaner.state.floatbank3_b_level	1	0.0
secondary_cleaner.state.floatbank5_a_level	1	0.0
primary_cleaner.state.floatbank8_b_level	1	0.0
rougher.state.floatbank10_a_level	1	0.0
primary_cleaner.state.floatbank8_d_level	1	0.0
primary_cleaner.state.floatbank8_c_level	1	0.0
rougher.state.floatbank10_a_air	1	0.0
primary_cleaner.state.floatbank8_a_level	1	0.0
rougher.state.floatbank10_c_level	1	0.0
rougher.state.floatbank10_c_air	1	0.0
rougher.output.tail_ag	1	0.0
rougher.state.floatbank10_b_air	1	0.0
rougher.state.floatbank10_b_level	1	0.0
final.output.concentrate_au	0	0.0
secondary_cleaner.output.tail_au	0	0.0
rougher.state.floatbank10_d_air	0	0.0
primary_cleaner.output.tail_au	0	0.0
rougher.state.floatbank10_e_level	0	0.0
final.output.recovery	0	0.0
final.output.tail_au	0	0.0
rougher.input.feed_ag	0	0.0
rougher.state.floatbank10_d_level	0	0.0
rougher.state.floatbank10_f_level	0	0.0
rougher.state.floatbank10_f_air	0	0.0
primary_cleaner.output.concentrate_ag	0	0.0
rougher.calculation.au_pb_ratio	0	0.0
rougher.output.concentrate_au	0	0.0
rougher.output.concentrate_ag	0	0.0
rougher.output.concentrate_pb	0	0.0
rougher.output.recovery	0	0.0
rougher.output.tail_au	0	0.0
rougher.output.tail_pb	0	0.0
rougher.output.tail_sol	0	0.0
primary_cleaner.input.feed_size	0	0.0
rougher.input.feed_au	0	0.0
primary_cleaner.output.concentrate_au	0	0.0
date	0	0.0

	count	mean	std	min	25%	50%	75%	max
rougher.input.feed_ag	19439.0	8.79	1.94	0.01	7.24	8.59	10.21	14.60
rougher.output.concentrate_ag	19439.0	11.99	2.73	0.00	10.63	12.10	13.74	24.48
primary_cleaner.output.concentrate_ag	19439.0	8.44	2.05	0.00	7.26	8.51	9.81	16.08
final.output.concentrate_ag	19438.0	5.17	1.37	0.00	4.25	5.07	5.90	16.00
rougher.output.tail_ag	19438.0	5.59	1.11	0.59	4.89	5.77	6.39	12.72
primary_cleaner.output.tail_ag	19435.0	16.15	3.55	0.00	14.02	15.84	18.01	29.46
secondary_cleaner.output.tail_ag	19437.0	14.59	4.27	0.00	12.64	15.64	17.36	23.26
final.output.tail_ag	19438.0	9.69	2.33	0.00	8.06	9.74	11.13	19.55

	0	1	2	3	4
rougher.input.feed_ag	0.010000	0.568441	1.015676	1.106758	1.116930
rougher.output.concentrate_ag	0.000000	0.010000	0.351874	0.905993	1.180162
primary_cleaner.output.concentrate_ag	0.000000	0.010000	0.018661	0.038086	0.116097
final.output.concentrate_ag	0.000000	0.010000	0.193950	0.289059	0.371783
rougher.output.tail_ag	0.594562	2.158369	2.170361	2.278989	2.316530
primary_cleaner.output.tail_ag	0.000000	0.235819	0.368401	0.437259	0.475679
secondary_cleaner.output.tail_ag	0.000000	0.017409	0.167573	0.248164	0.295360
final.output.tail_ag	0.000000	0.016790	0.078405	0.080674	0.101624

	count	mean	std	min	25%	50%	75%	max
rougher.input.feed_au	19439.0	8.27	1.96	0.01	6.85	8.13	9.77	13.73
rougher.output.concentrate_au	19439.0	19.77	3.75	0.00	18.80	20.28	21.72	28.82
primary_cleaner.output.concentrate_au	19439.0	32.12	5.63	0.00	30.46	32.84	35.05	45.93
final.output.concentrate_au	19439.0	44.08	5.13	0.00	43.40	45.01	46.28	52.76
rougher.output.tail_au	19439.0	1.82	0.68	0.02	1.40	1.81	2.21	9.69
primary_cleaner.output.tail_au	19439.0	3.91	1.59	0.00	2.92	3.62	4.59	17.79
secondary_cleaner.output.tail_au	19439.0	4.34	2.33	0.00	3.18	4.07	5.06	26.81
final.output.tail_au	19439.0	3.04	0.92	0.00	2.46	2.98	3.57	8.25

	0	1	2	3	4
rougher.input.feed_au	0.010000	0.568632	0.588152	0.625968	0.714036
rougher.output.concentrate_au	0.000000	0.010000	0.570819	1.889855	1.912582
primary_cleaner.output.concentrate_au	0.000000	0.010000	0.056590	0.135919	0.439876
final.output.concentrate_au	0.000000	0.010000	0.904254	1.063992	1.195223
rougher.output.tail_au	0.020676	0.024545	0.031234	0.031642	0.049731
primary_cleaner.output.tail_au	0.000000	0.045269	0.052055	0.056646	0.125645
secondary_cleaner.output.tail_au	0.000000	0.005877	0.033529	0.040478	0.042254
final.output.tail_au	0.000000	0.000777	0.005462	0.016073	0.020922

	count	mean	std	min	25%	50%	75%	max
rougher.input.feed_pb	19339.0	3.60	1.05	0.01	2.89	3.53	4.24	7.14
rougher.output.concentrate_pb	19439.0	7.61	1.80	0.00	6.73	7.74	8.57	13.62
primary_cleaner.output.concentrate_pb	19323.0	9.83	2.56	0.00	8.70	10.15	11.42	17.08
final.output.concentrate_pb	19438.0	9.98	1.67	0.00	9.14	10.10	11.04	17.03
rougher.output.tail_pb	19439.0	0.65	0.25	0.01	0.48	0.62	0.78	3.78
primary_cleaner.output.tail_pb	19418.0	3.44	1.39	0.00	2.50	3.27	4.18	9.63
secondary_cleaner.output.tail_pb	19427.0	5.78	2.77	0.00	3.86	5.34	7.79	17.04
final.output.tail_pb	19338.0	2.71	0.95	0.00	2.04	2.75	3.33	5.80

	0	1	2	3	4
rougher.input.feed_pb	0.010000	0.088552	0.125889	0.128727	0.157686
rougher.output.concentrate_pb	0.000000	0.010000	0.214360	0.629964	0.772184
primary_cleaner.output.concentrate_pb	0.000000	0.010000	0.021357	0.045857	0.090132
final.output.concentrate_pb	0.000000	0.010000	0.332391	0.571758	0.629302
rougher.output.tail_pb	0.007566	0.009014	0.010670	0.011073	0.013077
primary_cleaner.output.tail_pb	0.000000	0.049552	0.084681	0.089703	0.090518
secondary_cleaner.output.tail_pb	0.000000	0.007514	0.065370	0.073368	0.100373
final.output.tail_pb	0.000000	0.004508	0.018764	0.020475	0.030546

	count	mean	std	min	25%	50%	75%	max
rougher.input.feed_sol	19340.0	36.70	5.15	0.01	34.21	37.20	40.04	53.48
rougher.output.concentrate_sol	19416.0	28.81	5.94	0.00	27.22	29.84	32.19	38.35
primary_cleaner.output.concentrate_sol	19069.0	10.49	3.91	0.00	7.89	10.31	13.45	22.46
final.output.concentrate_sol	19228.0	9.50	2.79	0.00	7.72	9.22	10.95	19.62
rougher.output.tail_sol	19439.0	18.06	3.45	10.01	15.82	18.11	20.09	66.12
primary_cleaner.output.tail_sol	19377.0	7.97	2.21	0.00	6.75	8.18	9.55	22.28
secondary_cleaner.output.tail_sol	17691.0	7.17	3.92	0.00	3.85	7.69	10.41	26.00
final.output.tail_sol	19433.0	10.58	2.87	0.00	8.94	10.62	12.10	22.32

	0	1	2	3	4
rougher.input.feed_sol	0.010000	0.463867	1.393488	1.716285	3.091118
rougher.output.concentrate_sol	0.000000	0.010000	0.879593	1.715906	2.152051
primary_cleaner.output.concentrate_sol	0.000000	0.010000	0.016416	0.062800	0.067671
final.output.concentrate_sol	0.000000	0.010000	0.253231	0.268995	0.291099
rougher.output.tail_sol	10.014648	10.024005	10.035363	10.058485	10.059291
primary_cleaner.output.tail_sol	0.000000	0.033303	0.074203	0.126310	0.179470
secondary_cleaner.output.tail_sol	0.000000	0.014898	0.021719	0.024451	0.036183
final.output.tail_sol	0.000000	0.077731	0.104827	0.131452	0.160814

	count	mean	std	25%	50%	75%	max
rougher.input.floatbank10_sulfate	19405.0	12.33	3.43	10.00	12.00	14.72	36.12
rougher.input.floatbank10_xanthate	19431.0	6.02	1.13	5.40	6.01	6.80	9.70
rougher.input.floatbank11_sulfate	19395.0	12.06	3.78	10.00	12.00	14.64	37.98
rougher.input.floatbank11_xanthate	18986.0	6.07	1.10	5.40	6.10	6.80	9.70
primary_cleaner.input.depressant	19402.0	8.82	3.29	6.08	8.04	11.01	40.00
primary_cleaner.input.sulfate	19415.0	144.62	44.46	114.11	143.23	175.08	265.98
primary_cleaner.input.xanthate	19335.0	1.02	0.51	0.70	0.94	1.21	4.10

	0	1	2	3	4
rougher.input.floatbank10_sulfate	0.001083	0.001147	0.001164	0.001613	0.002683
rougher.input.floatbank10_xanthate	0.000568	0.001428	0.001822	0.001963	0.002150
rougher.input.floatbank11_sulfate	0.000077	0.000098	0.000163	0.000529	0.000726
rougher.input.floatbank11_xanthate	0.000290	0.000406	0.000578	0.000818	0.000953
primary_cleaner.input.depressant	0.000000	0.003839	0.007721	0.012244	0.012373
primary_cleaner.input.sulfate	0.003112	0.051279	0.067424	0.129139	0.396662
primary_cleaner.input.xanthate	0.000005	0.000154	0.000214	0.000291	0.000346

	0	1	2	3	4
rougher.input.floatbank10_sulfate	36.118275	31.393045	30.830567	30.631803	30.272362
rougher.input.floatbank10_xanthate	9.703448	9.674621	9.655247	9.517497	9.491715
rougher.input.floatbank11_sulfate	37.980648	32.780762	31.849375	30.011666	30.007997
rougher.input.floatbank11_xanthate	9.698407	9.668678	9.667279	9.514789	9.496275
primary_cleaner.input.depressant	40.000000	20.052255	20.046764	20.045827	20.038590
primary_cleaner.input.sulfate	265.983123	262.531502	262.293482	262.061026	262.003703
primary_cleaner.input.xanthate	4.102454	4.068384	4.027341	4.014188	4.012052

	count	mean	std	min	25%	50%	75%	max
rougher.input.feed_rate	19428.0	478.32	105.37	0.00	416.53	499.42	550.17	717.51
rougher.input.feed_size	19294.0	58.97	21.63	0.05	47.44	54.61	65.02	484.97
primary_cleaner.input.feed_size	19439.0	7.31	0.61	1.08	6.94	7.28	7.67	15.50

	0	1	2	3	4
rougher.input.feed_rate	0.002986	0.012786	0.022802	0.023393	0.023788
rougher.input.feed_size	0.046369	0.049194	0.052019	0.054843	9.659576
primary_cleaner.input.feed_size	1.080000	1.080007	4.970000	5.310000	5.410000

	count	mean	std	min	25%	50%	75%	max
rougher.state.floatbank10_a_air	19438.0	1105.32	160.83	-0.04	999.66	1001.07	1205.62	1521.98
rougher.state.floatbank10_b_air	19438.0	1320.22	176.29	-0.72	1199.63	1301.58	1448.63	1809.19
rougher.state.floatbank10_c_air	19438.0	1299.97	202.80	-0.06	1151.08	1300.21	1449.46	2499.13
rougher.state.floatbank10_d_air	19439.0	1211.56	204.02	-0.99	1061.05	1201.57	1352.88	1817.20
rougher.state.floatbank10_e_air	19003.0	1080.87	172.15	-1.98	998.68	1050.03	1199.43	1922.64
rougher.state.floatbank10_f_air	19439.0	1025.27	164.49	-2.59	901.00	999.92	1099.72	1706.31
primary_cleaner.state.floatbank8_a_air	19435.0	1589.35	129.00	0.00	1505.06	1600.60	1697.66	2103.10
primary_cleaner.state.floatbank8_b_air	19435.0	1591.34	131.81	0.00	1503.63	1600.52	1699.22	2114.91
primary_cleaner.state.floatbank8_c_air	19437.0	1586.67	136.11	0.00	1500.93	1600.25	1699.48	2013.16
primary_cleaner.state.floatbank8_d_air	19436.0	1542.91	246.20	0.00	1494.22	1599.45	1698.52	2398.90

	0	1	2	3	4
rougher.state.floatbank10_a_air	-4.206929e-02	-4.107792e-02	-0.020163	-0.020059	155.593194
rougher.state.floatbank10_b_air	-7.218064e-01	-6.619201e-01	-0.653384	-0.408072	270.696463
rougher.state.floatbank10_c_air	-5.663551e-02	-3.511233e-02	-0.026301	2.426204	265.315357
rougher.state.floatbank10_d_air	-9.922190e-01	-9.274013e-01	-0.554299	-0.124766	304.093664
rougher.state.floatbank10_e_air	-1.982295e+00	-1.970275e+00	-1.936877	-1.915434	94.372023
rougher.state.floatbank10_f_air	-2.593390e+00	-2.520792e+00	-2.430587	-2.328641	28.332355
primary_cleaner.state.floatbank8_a_air	5.445860e-32	7.805874e-32	0.006352	0.016993	304.788187
primary_cleaner.state.floatbank8_b_air	6.647490e-32	9.528241e-32	0.008700	0.023275	306.005110
primary_cleaner.state.floatbank8_c_air	4.033736e-32	5.781792e-32	0.015159	0.040552	1.421099
primary_cleaner.state.floatbank8_d_air	4.039220e-32	5.789652e-32	0.008655	0.023153	131.321672

	count	mean	std	min	25%	50%	75%	max
secondary_cleaner.state.floatbank2_a_air	19219.0	28.71	5.46	0.08	25.05	29.10	33.01	52.65
secondary_cleaner.state.floatbank2_b_air	19416.0	24.11	5.75	0.00	20.92	25.04	28.01	35.15
secondary_cleaner.state.floatbank3_a_air	19426.0	28.20	6.13	0.00	24.95	27.98	33.00	44.26
secondary_cleaner.state.floatbank3_b_air	19438.0	22.20	5.57	0.00	19.00	22.01	26.00	35.07
secondary_cleaner.state.floatbank4_a_air	19433.0	18.99	5.41	0.00	14.98	18.02	23.01	30.12
secondary_cleaner.state.floatbank4_b_air	19438.0	15.01	4.89	0.00	11.94	14.97	19.03	31.27
secondary_cleaner.state.floatbank5_a_air	19438.0	15.69	5.51	-0.37	10.99	15.00	18.03	43.71
secondary_cleaner.state.floatbank5_b_air	19438.0	12.20	5.33	0.53	8.97	11.02	14.02	27.93
secondary_cleaner.state.floatbank6_a_air	19437.0	18.96	5.55	-0.08	14.98	19.96	24.00	32.19

	0	1	2	3	4
secondary_cleaner.state.floatbank2_a_air	0.077503	1.220129e-01	1.425403e-01	1.440420e-01	1.627969e-01
secondary_cleaner.state.floatbank2_b_air	0.000000	6.778101e-18	8.210020e-18	9.309321e-18	1.276527e-17
secondary_cleaner.state.floatbank3_a_air	0.000000	1.833598e-21	1.957942e-17	2.495226e-17	3.309031e-17
secondary_cleaner.state.floatbank3_b_air	0.000000	6.137256e-22	1.521587e-17	1.158437e-16	3.130303e-10
secondary_cleaner.state.floatbank4_a_air	0.000000	1.079872e-16	2.706261e-10	7.271605e-05	7.949184e-04
secondary_cleaner.state.floatbank4_b_air	0.000000	2.489718e-17	8.306354e-17	2.030172e-10	6.802814e-05
secondary_cleaner.state.floatbank5_a_air	-0.372054	-3.719793e-01	-3.717388e-01	-3.714983e-01	-3.714907e-01
secondary_cleaner.state.floatbank5_b_air	0.528083	6.462077e-01	6.680429e-01	6.794086e-01	6.922236e-01
secondary_cleaner.state.floatbank6_a_air	-0.079426	1.681479e-02	1.273038e-01	1.507513e-01	1.603862e-01

	count	mean	std	min	25%	50%	75%	max
rougher.state.floatbank10_a_level	19438.0	-376.61	93.91	-657.95	-499.84	-300.26	-299.98	-273.78
rougher.state.floatbank10_b_level	19438.0	-467.25	54.67	-650.26	-500.20	-499.78	-400.62	-296.38
rougher.state.floatbank10_c_level	19438.0	-467.72	53.53	-647.54	-500.22	-499.71	-400.99	-292.16
rougher.state.floatbank10_d_level	19439.0	-468.24	53.92	-648.39	-500.39	-499.53	-401.64	-76.40
rougher.state.floatbank10_e_level	19439.0	-466.04	57.38	-649.44	-500.23	-499.66	-401.16	-139.75
rougher.state.floatbank10_f_level	19439.0	-466.27	57.85	-649.95	-500.52	-499.47	-401.51	-191.72
primary_cleaner.state.floatbank8_a_level	19438.0	-491.20	32.43	-798.64	-500.31	-499.93	-499.43	-57.20
primary_cleaner.state.floatbank8_b_level	19438.0	-492.19	33.79	-800.00	-500.44	-499.96	-499.37	-142.53
primary_cleaner.state.floatbank8_c_level	19438.0	-491.98	32.71	-799.96	-500.70	-499.91	-498.86	-150.94
primary_cleaner.state.floatbank8_d_level	19438.0	-488.02	43.96	-799.79	-500.46	-499.91	-499.06	-30.60

	0	1	2	3	4
rougher.state.floatbank10_a_level	-273.776101	-278.320576	-281.039299	-281.041967	-281.044634
rougher.state.floatbank10_b_level	-296.383080	-298.036622	-298.078576	-298.200051	-298.227537
rougher.state.floatbank10_c_level	-292.155509	-293.273988	-295.103390	-296.383919	-296.813767
rougher.state.floatbank10_d_level	-76.397327	-208.330673	-282.986203	-290.190398	-292.463560
rougher.state.floatbank10_e_level	-139.753025	-272.199598	-294.832453	-295.323395	-296.011250
rougher.state.floatbank10_f_level	-191.718480	-243.846376	-249.801043	-254.174195	-258.300939
primary_cleaner.state.floatbank8_a_level	-57.195404	-189.696878	-282.988443	-330.134203	-347.040563
primary_cleaner.state.floatbank8_b_level	-142.527229	-238.807102	-246.047833	-339.751525	-341.660086
primary_cleaner.state.floatbank8_c_level	-150.937035	-229.338520	-290.845404	-346.653450	-346.792216
primary_cleaner.state.floatbank8_d_level	-30.601828	-40.027176	-66.377169	-107.595099	-114.780485

	count	mean	std	min	25%	50%	75%	max
secondary_cleaner.state.floatbank2_a_level	19438.0	-502.37	53.57	-799.61	-500.23	-499.97	-499.67	-127.88
secondary_cleaner.state.floatbank2_b_level	19438.0	-503.23	56.57	-799.87	-500.25	-500.02	-499.79	-212.00
secondary_cleaner.state.floatbank3_a_level	19438.0	-486.20	54.41	-799.76	-500.21	-499.93	-499.33	-191.68
secondary_cleaner.state.floatbank3_b_level	19438.0	-493.74	43.37	-809.33	-500.12	-499.98	-499.81	-159.74
secondary_cleaner.state.floatbank4_a_level	19438.0	-485.89	47.76	-799.80	-500.70	-499.84	-498.25	-245.24
secondary_cleaner.state.floatbank4_b_level	19438.0	-461.08	67.41	-800.84	-500.19	-499.46	-400.12	-6.51
secondary_cleaner.state.floatbank5_a_level	19438.0	-488.68	34.53	-797.32	-500.46	-499.80	-498.38	-244.48
secondary_cleaner.state.floatbank5_b_level	19438.0	-487.15	38.35	-800.22	-500.13	-499.94	-499.44	-137.74
secondary_cleaner.state.floatbank6_a_level	19438.0	-505.44	37.69	-809.74	-500.73	-500.05	-499.50	-104.43

	0	1	2	3	4
secondary_cleaner.state.floatbank2_a_level	-127.875565	-166.933103	-172.639078	-179.583857	-194.455620
secondary_cleaner.state.floatbank2_b_level	-211.999273	-212.064665	-216.660814	-221.684802	-223.045323
secondary_cleaner.state.floatbank3_a_level	-191.678810	-246.396324	-246.479860	-252.878296	-262.843412
secondary_cleaner.state.floatbank3_b_level	-159.742331	-159.797516	-163.888539	-179.157914	-186.530007
secondary_cleaner.state.floatbank4_a_level	-245.239184	-326.000930	-334.928409	-351.922004	-366.990021
secondary_cleaner.state.floatbank4_b_level	-6.506986	-7.834746	-12.498243	-13.175555	-15.646201
secondary_cleaner.state.floatbank5_a_level	-244.483566	-275.073125	-282.534868	-287.910009	-288.963521
secondary_cleaner.state.floatbank5_b_level	-137.740004	-141.934397	-144.171219	-148.080088	-149.705075
secondary_cleaner.state.floatbank6_a_level	-104.427459	-123.307487	-136.679155	-183.202019	-194.978411

	count	mean	std	min	25%	50%	75%	max
rougher.calculation.floatbank10_sulfate_to_au_feed	19437.0	3393.052498	4943.845943	-486.598675	2611.425833	3018.353876	3676.769750	629638.977914
rougher.calculation.floatbank11_sulfate_to_au_feed	19437.0	3256.854171	5781.423659	-264.982785	2551.437587	2997.218088	3601.998911	718684.955855

Прогнозирование коэффициента извлечения золота из золотосодержащей руды

Содержание

Обзор данных¶

Импорт библиотек и предварительные настройки¶

Чтение файлов и загрузка данных¶

Общая характеристика набора данных¶

Понимание данных¶

Описание данных¶

Дата и время¶

Содержание металлов и солей в потоках¶

Содержание серебра в потоках¶

Содержание золота в потоках¶

Содержание свинца в потоках¶

Содержание солей других металлов в потоках¶

Вспомогательные вещества¶

Скорость потока и размер частиц¶

Объём воздуха¶

Уровень жидкости¶

Расчётные величины¶

Проверка правильности автоматического расчёта коэффициента извлечения¶

Предобработка данных¶

Обработка аномальных значений¶

Обработка пропусков¶

Исследовательский анализ данных¶

Изменение содержания металлов в основном потоке¶

Извлечение золота¶

Извлечение серебра¶

Извлечение свинца¶

Изменение суммарного содержания металлов в основном потоке¶

Изменение размеров гранул исходного сырья¶

Построение моделей¶

Разделение данных на выборки¶

Выделение признаков для формирования тестовой и тренировочной выборок¶

Выделение признаков, влияющих на прогноз значения коэффициента извлечения на стадии флотации¶

Функция расчёта метрики качества моделей¶

Обучение моделей¶

Решающее дерево¶

Случайный лес¶

Линейная регрессия¶

Оценка моделей¶

Оценка адекватности моделей¶

Общий вывод¶

Прогнозирование коэффициента извлечения золота
из золотосодержащей руды

	count	mean	std	min	25%	50%	75%	max
rougher.output.recovery	19439.0	83.328587	14.148086	0.0	81.037957	86.188892	90.010670	100.0
final.output.recovery	19439.0	67.050208	10.125840	0.0	63.299712	68.172738	72.686642	100.0

	auto_estimated_recovery	recalculated_recovery
13853	93.189832	93.189832
14275	89.600154	89.600154
3889	84.106063	84.106063
10087	81.037128	81.037128
288	80.930498	80.930498
13067	89.498303	89.498303
4105	86.456387	86.456387

Содержание золота в потоках, усл. ед.	В сырье	Со стадии флотации	Со стадии очистки 1	Во флотоконцентрате
кол-во значений	17555	17555	17555	17555
кол-во уникальных	17543	17544	17543	17543
мин.	0.885875	1.889855	0.05659	3.911529
-1.5IQR	3.673437	16.101152	26.09429	40.766088
25 %	6.82	18.93	30.57	43.42
мода	[7.495525360107423]	[18.65237236022949]	[30.35080909729004]	[46.10503005981445]
медиана	8.07	20.31	32.89	45.01
среднее ариф.	8.24	20.19	32.54	44.33
75 %	9.74	21.73	35.1	46.25
+1.5IQR	12.458585	24.522948	39.680395	49.254085
макс.	13.524339	28.824507	45.933934	51.571885
размах	12.638465	26.934652	45.877344	47.660355
межквартильный размах	2.928382	2.807266	4.528702	2.829332

Размер гранул сырья, усл. ед.	В тестовой выборке	В тренировочной выборке
кол-во значений	4626	12929
кол-во уникальных	4577	12852
мин.	9.659576	12.191554
-1.5IQR	27.230071	29.866792
25 %	47.66	47.32
мода	[55.46105787765386]	[55.1907258792174]
медиана	55.57	54.46
среднее ариф.	59.25	58.39
75 %	66.56	63.72
+1.5IQR	83.909566	79.061683
макс.	363.989699	484.967466
размах	354.330123	472.775913
межквартильный размах	18.893165	16.398297